Computer Sciences Commons^™

A Learning-To-Rank Based Fault Localization Approach Using Likely Invariants, Tien-Duy B. Le, David Lo, Claire Le Goues, Lars Grunske

Research Collection School Of Computing and Information Systems

Debugging is a costly process that consumes much of developer time and energy. To help reduce debugging effort, many studies have proposed various fault localization approaches. These approaches take as input a set of test cases (some failing, some passing) and produce a ranked list of program elements that are likely to be the root cause of the failures (i.e., failing test cases). In this work, we propose Savant, a new fault localization approach that employs a learning-to-rank strategy, using likely invariant diffs and suspiciousness scores as features, to rank methods based on their likelihood to be a root cause …

Practitioners' Expectations On Automated Fault Localization, Pavneet Singh Kochhar, Xin Xia, David Lo, Shanping Li

Research Collection School Of Computing and Information Systems

Software engineering practitioners often spend significant amount of time and effort to debug. To help practitioners perform this crucial task, hundreds of papers have proposed various fault localization techniques. Fault localization helps practitioners to find the location of a defect given its symptoms (e.g., program failures). These localization techniques have pinpointed the locations of bugs of various systems of diverse sizes, with varying degrees of success, and for various usage scenarios. Unfortunately, it is unclear whether practitioners appreciate this line of research. To fill this gap, we performed an empirical study by surveying 386 practitioners from more than 30 countries …

Cross-Modal Self-Taught Hashing For Large-Scale Image Retrieval, Liang Xie, Lei Zhu, Peng Pan, Yansheng Lu

Research Collection School Of Computing and Information Systems

Cross-modal hashing integrates the advantages of traditional cross-modal retrieval and hashing, it can solve large-scale cross-modal retrieval effectively and efficiently. However, existing cross-modal hashing methods rely on either labeled training data, or lack semantic analysis. In this paper, we propose Cross-Modal Self-Taught Hashing (CMSTH) for large-scale cross-modal and unimodal image retrieval. CMSTH can effectively capture the semantic correlation from unlabeled training data. Its learning process contains three steps: first we propose Hierarchical Multi-Modal Topic Learning (HMMTL) to detect multi-modal topics with semantic information. Then we use Robust Matrix Factorization (RMF) to transfer the multi-modal topics to hash codes which are …

Satisfiability Modulo Heap-Based Programs, Quang Loc Le, Jun Sun, Wei-Ngan Chin

Research Collection School Of Computing and Information Systems

In this work, we present a semi-decision procedure for a fragment of separation logic with user-defined predicates and Presburger arithmetic. To check the satisfiability of a formula, our procedure iteratively unfolds the formula and examines the derived disjuncts. In each iteration, it searches for a proof of either satisfiability or unsatisfiability. Our procedure is further enhanced with automatically inferred invariants as well as detection of cyclic proof. We also identify a syntactically restricted fragment of the logic for which our procedure is terminating and thus complete. This decidable fragment is relatively expressive as it can capture a range of sophisticated …

Metrics, Software Engineering, Small Systems – The Future Of Systems Development, William L. Honig

Computer Science: Faculty Publications and Other Works

In this talk I will introduce the importance of metrics, or measures, and the role they play in the development of high quality computer systems. I will review some key mega trends in computer science over the last three decades and then explain why I believe the trend to small networked systems, along with metrics and software engineering will define the future of high technology computer based systems.

I first learned about metrics at the Bell System where everything was measured. Metrics can be understood easily if you think of them as measures, for example of calories or salt in …

An Interference-Free Programming Model For Network Objects, Mischael Schill, Christopher M. Poskitt, Bertrand Meyer

Research Collection School Of Computing and Information Systems

Network objects are a simple and natural abstraction for distributed object-oriented programming. Languages that support network objects, however, often leave synchronization to the user, along with its associated pitfalls, such as data races and the possibility of failure. In this paper, we present D-Scoop, a distributed programming model that allows for interference-free and transaction-like reasoning on (potentially multiple) network objects, with synchronization handled automatically, and network failures managed by a compensation mechanism. We achieve this by leveraging the runtime semantics of a multi-threaded object-oriented concurrency model, directly generalizing it with a message-based protocol for efficiently coordinating remote objects. We present …

Why A Testing Career Is Not The First Choice Of Engineers, Pradeep Kashinath Waychal, Luiz Fernando Capretz

Electrical and Computer Engineering Publications

No abstract provided.

Demo: Multi-Device Gestural Interfaces, Tran Huy Vu, Youngki Lee, Archan Misra

Research Collection School Of Computing and Information Systems

Varieties of wearable devices such as smart watches, Virtual/Augmented Reality devices (AR/VR) are much more affordable with interesting capabilities. In our vision, a person may use more than one devices at a time, and they form an eco-system of wearable devices. Therefore, we aim to build a system where an application expands its input and output among different devices, and adapts its input/output stream for different contexts.

Poster: Sonicnect: Accurate Hands-Free Gesture Input System With Smart Acoustic Sensing, Maotian Chang, Ping Li, Panlong Yang, Jie Xiong, Chang Tian

Research Collection School Of Computing and Information Systems

This work presents Sonicnect, an acoustic sensing system with smartphone that enables accurate hands-free gesture input. Sonicnect leverages the embedded microphone in the smartphone to capture the subtle audio signals generated with fingers touching on the table. It supports 9 commonly used gestures (click, flip, scroll and zoom, etc) with above 92% recognition accuracy, and the minimum gesture movement could be 2cm. Distinguishable features are then extracted by exploiting spatio-temporal and frequency properties of the subtle audio signals. We conduct extensive real environment experiments to evaluate its performance. The results validate the effectiveness and robustness of Sonicnect.

Demo: Wearable Application To Manage Problem Behavior In Children With Neurodevelopmental Disorders, Camellia Zakaria, Richard C. Davis

Research Collection School Of Computing and Information Systems

Managing problem behaviors in children with neurodevelopmental disorders can be challenging. Such behaviors may discourage social participation and learning. Many of these behaviors warrant intervention, however, are challenging for caregivers to constantly supervise. Previous work focused on developing recognition systems for stereotypical and aggressive behaviors. Researchers also developed visualization interface for caregivers to better understand their child’s needs. Our goal however, is to design an independent behavior management application to help children manage problem behaviors with minimal supervision.We conducted a field study at a school for children with special needs in Singapore, and interviewed ten teachers. This study helped us …

Automating Self Evaluations For Software Engineers, Jonathan Rodrigo A. Miranda

Master's Theses

Software engineers frequently compose self-evaluations as part of employee perfor- mance reviews. These evaluations can be a key artifact for assessing a software engineer’s contributions to a team and organization, and for generating useful feed- back. Self-evaluations can be challenging because a) they can be time consuming, b) individuals may forget about important contributions especially when the review period is long such as a full year, c) some individuals can consciously or unconsciously overstate their contributions, and d) some individuals can be reluctant to describe their contributions for fear of appearing too proud [24].

UNBIASED, Useful New Basic Interactive Automated …

Fusing Wifi And Video Sensing For Accurate Group Detection In Indoor Spaces, Kasthuri Jayarajah, Zaman Lantra, Archan Misra

Research Collection School Of Computing and Information Systems

Understanding one's group context in indoor spaces is useful for many reasons - e.g., at a shopping mall, knowing a customer's group context can help in offering context-specific incentives, or estimating taxi demand for customers exiting the mall. Group detection and monitoring using WiFi-based indoor location traces fails when users are invisible (either because they don't carry smartphones, or because their WiFi is turned OFF) or when location tracking is inaccurate. In this paper, we propose a multi-modal group detection system that fuses two independent modes: video and WiFi, for detecting groups with low latency and high accuracy. We present …

The Elder Scrolls V: Skyrim Stamina Combat Overhaul, Richard Rattner

Liberal Arts and Engineering Studies

No abstract provided.

Demo: Drumming Application Using Commodity Wearable Devices, Bharat Dwivedi, Archan Misra, Youngki Lee

Research Collection School Of Computing and Information Systems

We aim to develop a drumming application in which individual can play drums using multiple wearable and mobile devices. Our vision is to tap out different rythms in the air using smart watches as a virtual drum stick and smart phone would act as a drum kit. Same user interface can be visualized in smart glasses. Here, our prime target is to use multiple commodity wearable devices (non-commodity i.e. Myo arm band) and smart phones for recognizing new (or same type of here) types of multi limb gestural context and building an adaptive application interface and allow such gesture recognition …

Livelabs: Building In-Situ Mobile Sensing And Behavioural Experimentation Testbeds, Kasthuri Jayarajah, Rajesh Krishna Balan, Meera Radhakrishnan, Archan Misra, Youngki Lee

Research Collection School Of Computing and Information Systems

In this paper, we present LiveLabs, a first-of-its-kind testbed that isdeployed across a university campus, convention centre, and resortisland and collects real-time attributes such as location, group contextetc., from hundreds of opt-in participants. These venues, data,and participants are then made available for running rich humancentricbehavioural experiments that could test new mobile sensinginfrastructure, applications, analytics, or more social-sciencetype hypotheses that influence and then observe actual user behaviour.We share case studies of how researchers from aroundthe world have and are using LiveLabs, and our experiences andlessons learned from building, maintaining, and expanding LiveLabsover the last three years.

Seeking Independent Management Of Problem Behavior: A Proof-Of-Concept Study With Children And Their Teachers, Camellia Zakaria, Richard C. Davis, Zachary Walker

Research Collection School Of Computing and Information Systems

Problem behaviors are particularly common in children with neurodevelopmental disorders like Autism and Down syndrome. These behaviors sometimes discourage social inclusion, inhibit learning development, and cause severe injuries, but caregivers are often unable to attend to their children immediately when the behaviors occur. Recent research shows that problem behavior can be automatically detected with wearable devices, but it is still not clear how to reduce caregivers' burdens and facilitate academic, social, and functional development of children with problem behaviors. We conducted a field study at a school with 21 children who exhibit problem behaviors and found that they needed frequent …

Poster: Android Whole-System Control Flow Analysis For Accurate Application Behavior Modeling, Huu Hoang Nguyen

Research Collection School Of Computing and Information Systems

Android, the modern operating system for smartphones, together with its millions of apps, has become an important part of human life. There are many challenges to analyzing them. It is important to model the mobile systems in order to analyze the behaviors of apps accurately. These apps are built on top of interactions with Android systems. We aim to automatically build abstract models of the mobile systems and thus automate the analysis of mobile applications and detect potential issues (e.g., leaking private data, causing unexpected crashes, etc.). The expected results will be the accuracy models of actual various versions of …

Demo: Ta$Ker: Campus-Scale Mobile Crowd-Tasking Platform, Nikita Jaiman, Thivya Kandappu, Randy Tandriansyah, Archan Misra

Research Collection School Of Computing and Information Systems

We design and develop TA$Ker, a real-world mobile crowd- sourcing platform to empirically study the worker responses to various task recommendation and selection strategies.

Condensing Class Diagrams With Minimal Manual Labeling Cost, Xinli Yang, David Lo, Xin Xia, Jianling Sun

Research Collection School Of Computing and Information Systems

Traditionally, to better understand the design of a project, developers can reconstruct a class diagram from source code using a reverse engineering technique. However, the raw diagram is often perplexing because there are too many classes in it. Condensing the reverse engineered class diagram into a compact class diagram which contains only the important classes would enhance the understandability of the corresponding project. A number of recent works have proposed several supervised machine learning solutions that can be used for condensing reverse engineered class diagrams given a set of classes that are manually labeled as important or not. However, a …

Automated Identification Of High Impact Bug Reports Leveraging Imbalanced Learning Strategies, Xinli Yang, David Lo, Qiao Huang, Xin Xia, Jianling Sun

Research Collection School Of Computing and Information Systems

In practice, some bugs have more impact than others and thus deserve more immediate attention. Due to tight schedule and limited human resource, developers may not have enough time to inspect all bugs. Thus, they often concentrate on bugs that are highly impactful. In the literature, high impact bugs are used to refer to the bugs which appear in unexpected time or locations and bring more unexpected effects, or break pre-existing functionalities and destroy the user experience. Unfortunately, identifying high impact bugs from the thousands of bug reports in a bug tracking system is not an easy feat. Thus, an …

Cace: Exploiting Behavioral Interactions For Improved Activity Recognition In Multi-Inhabitant Smart Homes, Mohammad Arif Ul Alam, Nirmalya Roy, Archan Misra, Joseph Taylor

Research Collection School Of Computing and Information Systems

We propose CACE (Constraints And Correlations mining Engine) which investigates the challenges of improving the recognition of complex daily activities in multi-inhabitant smart homes, by better exploiting the spatiotemporal relationships across the activities of different individuals. We first propose and develop a loosely-coupled Hierarchical Dynamic Bayesian Network (HDBN), which both (a) captures the hierarchical inference of complex (macro-activity) contexts from lower-layer microactivity context (postural and improved oral gestural context), and (b) embeds the various types of behavioral correlations and constraints (at both micro-and macro-activity contexts) across the individuals. While this model is rich in terms of accuracy, it is computationally …

How Long Will This Live? Discovering The Lifespans Of Software Engineering Ideas, Subhajit Datta, Santonu Sarkar, A. S. M Sajeev

Research Collection School Of Computing and Information Systems

We all want to be associated with long lasting ideas; as originators, or at least, expositors. For a tyro researcher or a seasoned veteran, knowing how long an idea will remain interesting in the community is critical in choosing and pursuing research threads. In the physical sciences, the notion of half-life is often evoked to quantify decaying intensity. In this paper, we study a corpus of 19,000+ papers written by 21,000+ authors across 16 software engineering publication venues from 1975 to 2010, to empirically determine the half-life of software engineering research topics. In the absence of any consistent and well-accepted …

Deepsense: A Gpu-Based Deep Convolutional Neural Network Framework On Commodity Mobile Devices, Huynh Nguyen Loc, Rajesh Krishna Balan, Youngki Lee

Research Collection School Of Computing and Information Systems

Recently, a branch of machine learning algorithms called deep learning gained huge attention to boost up accuracy of a variety of sensing applications. However, execution of deep learning algorithm such as convolutional neural network on mobile processor is non-trivial due to intensive computational requirements. In this paper, we present our early design of DeepSense - a mobile GPU-based deep convolutional neural network (CNN) framework. For its design, we first explored the differences between server-class and mobile-class GPUs, and studied effectiveness of various optimization strategies such as branch divergence elimination and memory vectorization. Our results show that DeepSense is able to …

Demo: Gpu-Based Image Recognition And Object Detection On Commodity Mobile Devices, Loc Nguyen Huynh, Rajesh Krishna Balan, Youngki Lee

Research Collection School Of Computing and Information Systems

In this demo, we show that it is feasible to execute CNN for vision sensing tasks directly on mobile devices by leveraging integrated GPU. We propose our design of DeepSense framework based on OpenCL to execute deep learning algorithms in energy-efficient and fast manner.

Jasper: Sensing Gamers' Emotions Using Physiological Sensors, Sinh Huynh, Youngki Lee, Taiwoo Park, Rajesh Krishna Balan

Research Collection School Of Computing and Information Systems

This paper aims to develop a system that evaluates the emotional experience of gamers based on physiological changes. A within-subject experiment with 22 participants has been designed to investigate the effects of difficulty level and social playing mode on player emotions and to examine the correlation between each emotion and the physiological changes. We demonstrate the feasibility of using commodity wearable physiological sensing devices to recognize mobile gamer's emotion. Specifically, our system performs 3-level excitement classification at an accuracy of 77.38% and binary classification of happiness state at an accuracy of 73.21%. These classification results show the potential of using …

Small Scale Deployment Of Seat Occupancy Detectors, Nguyen Huy Hoang Huy, Gihan Hettiarachchi, Youngki Lee, Rajesh Krishna Balan

Research Collection School Of Computing and Information Systems

In this paper, we present the results of a small-scale field deployment of our capacitance-based seat occupancy detector. We deployed our sensors to 36 seats in our university library and measured the performance of our system over a period of 8 weeks. As part of this deployment, we had to tackle numerous real-world deployment issues such as hardware failure, variations in signal quality, and interference caused by multiple objects in near proximity. We present our overall system design, along with the modifications we made to tackle various real-world problems. Finally, we present the results of our deployment which showed that …

Poster: A Device-Free Evaluation System For Gymnastics Using Passive Rfid Tags, Binbin Xie, Jie Xiong, Dingyi Fang, Xiaojiang Chen, Anwen Wang, Zhanyong Tang

Research Collection School Of Computing and Information Systems

No abstract provided.

Efficient Multi-Class Selective Sampling On Graphs, Peng Yang, Peilin Zhao, Zhen Hai, Wei Liu, Hoi, Steven C. H., Xiao-Li Li

Research Collection School Of Computing and Information Systems

A graph-based multi-class classification problem is typically converted into a collection of binary classification tasks via the one-vs.-all strategy, and then tackled by applying proper binary classification algorithms. Unlike the one-vs.-all strategy, we suggest a unified framework which operates directly on the multi-class problem without reducing it to a collection of binary tasks. Moreover, this framework makes active learning practically feasible for multi-class problems, while the one-vs.-all strategy cannot. Specifically, we employ a novel randomized query technique to prioritize the informative instances. This query technique based on the hybrid criterion of "margin" and "uncertainty" can achieve a comparable mistake bound …

Poster: Air Quality Friendly Route Recommendation System, Savina Singla, Divya Bansal, Archan Misra

Research Collection School Of Computing and Information Systems

To model the overall personal inhalation of hazardous gases through the air (both indoor and outdoor) by an individual, provide air quality friendly route recommendations, thus raising the overall quality of urban movement and living healthy life.

Demo: Smartwatch Based Shopping Gesture Recognition, Meeralakshmi Radhakrishnan, Sharanya Eswaran, Sougata Sen, Vigneshwaran Subbaraju, Archan Misra, Rajesh Krishna Balan

Research Collection School Of Computing and Information Systems

In the current retail segment, the retail store owners are keen to understand the browsing behavior and purchase pattern of the shoppers inside the physical stores. Profiling the behavior of the shopper is key to success for any marketing strategies that can optimize or personalize shopping-related services in real-time. We envision that exploiting the knowledge of real-time behavior of shopper’s in-store activities enables novel applications such as: (a) targeted advertising or recommendations: based on longer term shopper profiles, (b) proactive retail help to assist the shoppers who are confused in choosing between two items, (c) smart reminders that can remind …